Agent skill

project-orchestrator:verification

Use before claiming ANY work is complete — ensures evidence-based verification across all affected services, correct git state, and deployment confirmation.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/verification-gigafyde-project-orchestrator

SKILL.md

Verification Before Completion

Overview

Claiming work is complete without verification is dishonesty, not efficiency.

Core principle: Evidence before claims, always.

Violating the letter of this rule is violating the spirit of this rule.

Config Loading

  1. Check if .project-orchestrator/project.yml exists
  2. If yes: parse and extract services (test commands, auto_deploy), plans_dir, plans_structure
  3. If no: use defaults (monorepo, root, auto-detect test command)

The Iron Law

NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE

If you haven't run the verification command in this message, you cannot claim it passes.

The Gate Function

BEFORE claiming any status or expressing satisfaction:

1. IDENTIFY: What command proves this claim?
2. RUN: Execute the FULL command (fresh, complete)
3. READ: Full output, check exit code, count failures
4. VERIFY: Does output confirm the claim?
   - If NO: State actual status with evidence
   - If YES: State claim WITH evidence
5. ONLY THEN: Make the claim

Skip any step = lying, not verifying

Verification Checklist

1. Run affected service tests

Every affected service MUST have its tests run fresh.

Test commands come from config.services[name].test, the project's root CLAUDE.md, or auto-detection.

Multi-service rule: When a change spans services, verify ALL affected services — not just the one you changed.

2. Cross-service contract verification

If API shapes changed between services:

Check How
Producer response matches contract Read the controller/endpoint, verify shape
Consumer expects correct shape Read the API call site, verify it parses correctly
Event payload matches Verify publisher and consumer agree on schema

3. Git state verification

Verify each affected service has clean git state:

bash
# Verify correct service repo and branch
cd <service> && git branch --show-current

# Verify changes are committed
cd <service> && git status

# Verify no unrelated files staged
cd <service> && git diff --staged --name-only

4. Deployment verification (auto-deploy services)

For services where config.services[name].auto_deploy is true:

  1. Wait for deployment to complete
  2. Ask user to confirm deployment success (check logs, verify endpoint)
  3. Do NOT claim "deployed and working" without user confirmation

5. Living state doc update

If working from a plan in {config.plans_dir}/:

  • Update task status in Implementation Tasks table
  • Log verification results in Implementation Log
  • Update overall status if all tasks complete
  • When status becomes complete:
    • If config.plans_structure is standard: move doc to {plans_dir}/completed/ and update {plans_dir}/INDEX.md
    • If flat: just update status in the doc

Common Failures

Claim Requires Not Sufficient
"Tests pass" Test command output: 0 failures Previous run, "should pass"
"Build succeeds" Build command: exit 0 "Linter passed"
"Bug fixed" Test original symptom: passes "Code changed, assumed fixed"
"All services verified" Test output from EACH service Testing only one of three affected
"Deployed and working" User confirmed endpoint works "Push succeeded"
"Committed to right branch" git branch output Assumption
"No regressions" Full test suite: 0 failures Spot-checking one test

Red Flags — STOP

If you catch yourself:

  • Using "should", "probably", "seems to"
  • Expressing satisfaction before verification ("Great!", "Done!")
  • About to commit/push without running tests
  • Trusting a teammate agent's success report without checking
  • Only testing the service you changed, not consumers/producers
  • Saying "deployed" without user confirming the endpoint

ALL of these mean: STOP. Run the verification.

Common Rationalizations

Excuse Reality
"Should work now" RUN the verification
"I'm confident" Confidence ≠ evidence
"Linter passed" Linter ≠ tests ≠ build
"Agent said success" Verify independently
"Only changed one file" One file can break everything
"Tests passed earlier" Earlier ≠ now. Run again.
"Other service wasn't touched" Contract changes affect consumers too

Quick Reference

Phase What to Verify How
Tests Each affected service Run config.services[name].test or auto-detected command
Contracts Producer + consumer match Read both sides, compare shapes
Git Correct repo, branch, clean state git branch, git status per service
Deploy Endpoint works User confirms after push
State doc Updated with results Edit {plans_dir}/*.md, move to completed/ when done (standard structure)

Related Commands & Skills

When Action
Verification passed, ready to merge Suggest user run /project:finish
Changelog needed Suggest user run /project:changelog

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results